audio traces on a pcb?

[justDIY]

is there anything special I need to do, in regards to laying down traces that will be carrying 'line level' audio signals?

I'm workin up a pcb for an audio switcher (router?) ... it will take four stereo inputs, and using 74hc4052's, route that audio to one of four outputs. so I have four 4052's, each sharing a connection to the four inputs

my main concern is crosstalk between the channels ... I plan to use a hardware mute, where the output of the 4052 is turned off if there's no signal selected, but incase there's some dead air in the input signal, I don't want to be hearing a neighboring channel. I have laid down as much copper as possible (is that called a pour?) on the board, both top and bottom layers, and I've kept fairly generous spacing between the traces

my second concern, what should I do for ESD protection ... in the past with DC circuits, I've used a pair of reverse biased schottky diodes, one to vcc and one to gnd, but that trick won't work with an AC audio signal will it?

third question ... how often should I join the top and bottom ground planes, or should I try to minimize the joining? I put in a few vias, in areas where the plane has covered some distance without any through holes.

The 4052's are controlled by a pair of 595 registers, and I'll either feed them from the PC serial port or a microcontroller

any feedback is appericated. I haven't gone through and mitered the corners yet, and I noticed I have a few tracks overlapping on different levels, when they don't have to... I'm also short on decoupling caps on the Vee rail ... just plain ran out of room, need to get a bigger shoe horn.

**broken link removed**

 

[joecool85]

There is really no difference in audio or power traces on PCBs, the only thing to keep in mind when designing is that ideally its better to keep the signal traces away from the power traces or sometimes you get interference.

 

[mstechca]

...ideally its better to keep the signal traces away from the power traces or sometimes you get interference.

This is recommended because two long tracks that are too close together could create a high capacitance. Remember that a capacitor is made of an insulator sandwiched between two conductors.

In a radio, a capacitor can affect the receiving frequency.

I would advise not having two long tracks in parallel, unless you have a capacitor between those two tracks and a higher capacitor won't hurt. For example, the +ve and ground tracks can be close together, since they help increase the decoupling capacitor.

how often should I join the top and bottom ground planes, or should I try to minimize the joining?

It depends on how much time and money you want to put in.

I have tried a double-sided board once before, but since I don't have professional tools, I have to revert to single-sided boards. Also, single sided boards are cheaper.

My answer for you is only use vias if it is 100% necessary. and it is if you find that you are running too many tracks too close together on the board. I wouldn't be surprised, considering the way IC's are made now a days.

 

[Russlk]

Line level (1 volt RMS) can be carried on the narrowest line you can make. There should not be any problem in cross talk at audio frequencies, especially if the lines are terminated in 600 ohms somewhere. The audio cannot exceed VCC & VEE and still pass thru the switches, so the protection diodes will still work.

 

[mstechca]

before taking the narrow track advice to heart, you need to consider that the narrower your track is the higher resistace the track is. Likewise, the longer the track, the higher the resistance as well. Also, narrow tracks tend to be easier to break than larger tracks because there is less copper bonded to the board.

The resistance might not have such a big effect in digital electronics, but in analog electronics (especially inside a radio circuit), a few centimeters can make a big difference.

 

[justDIY]

here is the latest revision

I have added 100nf decoupling caps to the postive and negative rails of all the chips now, and tried to locate the cap as close to the chip as possible.

The two long parallel tracks are digitial signals, running from outputs on the 595 to the /E inputs on two of the 4052 - a slight capactiance on these lines should be acceptable, since they'll be driven high or low, and I don't expect or require ultra fast switching speed ... this is my 'hardware' mute option

Some leds were added for that 'eye candy' factor, as well as extending some un-used ports from the 595 to a 'general purpose output' connector, with current limiting resistors inline, just to protect the 595 incase of a external short. I think I'll add some 5.1v zeners (grounded) to the GP outputs, so any stray voltages coming in over 5 volts should get shunted to the ground?

I'm sending these boards out to be fabbed - this deff. exceeds my skill level for fabrication.

Glad to hear that the simple dual diode esd trick still works... I guess I forgot that I just connect the 'low' diode to vee and not to gnd - duh... I'll shoe horn some schottkys in on the inputs and outputs

**broken link removed**

schematic here.

 

[Nigel Goodwin]

before taking the narrow track advice to heart, you need to consider that the narrower your track is the higher resistace the track is. Likewise, the longer the track, the higher the resistance as well. Also, narrow tracks tend to be easier to break than larger tracks because there is less copper bonded to the board.

The resistance might not have such a big effect in digital electronics, but in analog electronics (especially inside a radio circuit), a few centimeters can make a big difference.

You're talking COMPLETE RUBBISH again, there's very, very, rarely any problems with the resistance of the copper tracks, and only under very specific (blindingly obvious) circumstances.

Where you've got it wrong is that for VHF (and above) circuits layout is critical, and PCB tracks have significant inductance and capacitance at those frequencies - with both inductors and capacitors both actually constructed out of the PCB material quite commonly.

 

[mstechca]

before taking the narrow track advice to heart, you need to consider that the narrower your track is the higher resistace the track is. Likewise, the longer the track, the higher the resistance as well. Also, narrow tracks tend to be easier to break than larger tracks because there is less copper bonded to the board.

The resistance might not have such a big effect in digital electronics, but in analog electronics (especially inside a radio circuit), a few centimeters can make a big difference.

there's very, very, rarely any problems with the resistance of the copper tracks, and only under very specific (blindingly obvious) circumstances.

I was telling him information that should be considered, should he decide to use extra-long tracks.

...for VHF (and above) circuits layout is critical, and PCB tracks have significant inductance

According to DC conditions, a resistor and an inductor are almost equal. In AC, they are completely different.

 

[Nigel Goodwin]

I was telling him information that should be considered, should he decide to use extra-long tracks.

I didn't realise you could buy PCB by the mile? :lol:

...for VHF (and above) circuits layout is critical, and PCB tracks have significant inductance

According to DC conditions, a resistor and an inductor are almost equal. In AC, they are completely different.

Again, very little resemblance to the way things are! - resistors and inductors are NOT the same at DC.

 

[mstechca]

resistors and inductors are NOT the same at DC.

Why not?

They both have resistance.

 

[Nigel Goodwin]

resistors and inductors are NOT the same at DC.

Why not?

They both have resistance.

Because an inductor is more like a piece of wire than a resistor at DC, in fact it IS just a piece of wire at DC :lol:

A very low value resistor is similar to an inductor, and often will consist of wire round around a core, actually making an inductor.

But this is only in a very small minority of resistors, your sweeping statement doesn't apply in 99.9999....% of cases.

 

[Oznog]

Nigel's saying that for audio frequencies the inductance and capacitance of the traces is not significant, unless perhaps something really weird is happening like you made a 3 ft PCB. Likewise the resistance is not significant with audio signal loads. If this were a power amplifier output for a many-amps subwoofer, perhaps that could change. But a 1Vpp audio signal into a 600 ohm load is only 1mA RMS. A 20mOhm trace would drop 1/30,000th of the voltage.

However I am not clear Nigel when you allude to "board layout is critical". While this addage is certainly true, which features do you see in his layout that you question?

 

[Nigel Goodwin]

However I am not clear Nigel when you allude to "board layout is critical". While this addage is certainly true, which features do you see in his layout that you question?

None, you've only quoted part of the sentence

for VHF (and above) circuits layout is critical

which was in answer to mstechca.